Question about discretize()
Question about discretize()
I am currently trying to use the discretize() method to discretize my data set. What do the numbers in the double[] represent? How do I convert a non-discretized node into a discretized node?
-
- Site Admin
- Posts: 1417
- Joined: Mon Nov 26, 2007 5:51 pm
Re: Question about discretize()
These are the interval boundaries.What do the numbers in the double[] represent?
Node as in node in the network or rather a variable in the dataset?How do I convert a non-discretized node into a discretized node?
Re: Question about discretize()
I think variable in the dataset, here I am trying to convert file file1 (as a csv) intoshooltz wrote:These are the interval boundaries.What do the numbers in the double[] represent?
That is what I was thinking, but the gui discretize boundries through the program doesn't match with the results of this.
Node as in node in the network or rather a variable in the dataset?How do I convert a non-discretized node into a discretized node?
file2 (as a csv).
-
- Site Admin
- Posts: 1417
- Joined: Mon Nov 26, 2007 5:51 pm
Re: Question about discretize()
Are you sure you're using the same discretization algorithm and bin count?That is what I was thinking, but the gui discretize boundries through the program doesn't match with the results of this.
Re: Question about discretize()
I figured out that the mismatch in data was due to the fact that when I used the GUI I highlighted all the variables and discretized them all at the same time. I assume this led the program to discretize them all together rather then separately as I thought it would work. Is there a way to discretize multiple variables together using the discretize method?
-
- Site Admin
- Posts: 1417
- Joined: Mon Nov 26, 2007 5:51 pm
Re: Question about discretize()
No. You'll need to invoke discretize() for each variable separately.Tiranous wrote: Is there a way to discretize multiple variables together using the discretize method?